Saudi Journal of Biological Sciences 27 (2020) 3514–3528

Contents lists available at ScienceDirect

Saudi Journal of Biological Sciences

journal homepage: www.sciencedirect.com

Original article

Comparative transcriptome profiling of pomegranate genotypes having

resistance and susceptible reaction to Xanthomonas axonopodis pv.
punicae
Nripendra Vikram Singh a,⇑, Shilpa Parashuram a, Jyotsana Sharma a, Roopa Sowjanya Potlannagari a,
Dhinesh Babu Karuppannan a, Ram Krishna Pal a, Prakash Patil a, Dhananjay M. Mundewadikar a,
Vipul R. Sangnure a, P.V. Parvati Sai Arun b, Naresh V.R. Mutha c, Bipin Kumar c, Abhishek Tripathi c,
Sathish Kumar Peddamma c, Harish Kothandaraman c, Sailu Yellaboina c, Dushyant Singh Baghel c,
Umesh K. Reddy d
a
ICAR-National Research Centre on Pomegranate, Solapur, Maharashtra 413255, India
b
Chaitanya Bharati Institute of Technology, Hyderabad, Telangana 50075, India
c
Nucleome Informatics Private Limited., Hyderabad, Telangana State 500049, India
d
Gus R. Douglass Institute and Department of Biology, West Virginia State University, Institute, WV, USA

a r t i c l e i n f o a b s t r a c t

Article history: Pomegranate (Punica granatum L.) is an important fruit crop, rich in fiber, vitamins, antioxidants, minerals
Received 22 March 2020 and source of different biologically active compounds. The bacterial blight caused by Xanthomonas axo-
Revised 16 July 2020 nopodispv. punicae is a serious threat to the crop leading to 60–80% yield loss under epiphytotic condi-
Accepted 18 July 2020
tions. In this work, we have generated comparative transcriptome profile to mark the gene expression
Available online 23 July 2020
signatures during resistance and susceptible interactions. We analyzed leaf and fruits samples of moder-
ately resistant genotype (IC 524207) and susceptible variety (Bhagawa) of pomegranate at three progres-
Keywords:
sive infection stages upon inoculation with the pathogen. RNA-Seq with the Illumina HiSeq 2500
Pomegranate
Bhagawa
platform revealed 1,88,337 non-redundant (nr) transcript sequences from raw sequencing data, for a
Moderately resistant total of 34,626 unigenes with size >2 kb. Moreover, 85.3% unigenes were annotated in at least one of
IC 524207 the seven databases examined. Comparative analysis of gene-expression signatures in resistant and sus-
Transcriptome ceptible varieties showed that the genes known to be involved in defense mechanism in plants were up-
Bacterial blight regulated in resistant variety. Gene Ontology (GO) analysis successfully annotated 90,485 pomegranate
Xanthomonas unigenes, of which 68,464 were assigned to biological, 78,107 unigenes molecular function and 44,414
to cellular components. Significantly enriched GO terms in DEGs were related to oxidations reduction
biological process, protein binding and oxidoreductase activity. This transcriptome data on pomegranate
could help in understanding resistance and susceptibility nature of cultivars and further detailed fine
mapping and functional validation of identified candidate gene would provide scope for resistance breed-
ing programme in pomegranate.
Ó 2020 The Authors. Published by Elsevier B.V. on behalf of King Saud University. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Pomegranate (Punica granatum L.) is an important fruit crop in

countries such as India, Iran, China, and Turkey (Soloklui et al.,
⇑ Corresponding author.
2012). It is a perennial fruit tree that is native from Iran to the
E-mail address: nripendras72@gmail.com (N.V. Singh).
Himalayan Mountains in northern India (Morton, 1987, Soloklui
Peer review under responsibility of King Saud University.
et al., 2012). India and Iran collectively produce more than half
the world’s pomegranate (Petersen et al., 2010, Sharma et al.,
2010). Pomegranate aril, seed, rind, flower, bark and root produce
different types of biologically active compounds and phytochemi-
Production and hosting by Elsevier
cals, including gallotannins, ellagic acid, flavonoids, antioxidants,

https://doi.org/10.1016/j.sjbs.2020.07.023
1319-562X/Ó 2020 The Authors. Published by Elsevier B.V. on behalf of King Saud University.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528
terpenoids and alkaloids, and are used in treating diseases such as
atherosclerosis, breast cancer, skin cancer, and prostate cancer
(Bayazit and Caliskan, 2018, MayuoniKirshinbaum and Porat,
2014, Syed et al., 2013, Ophir et al., 2014). With the characteristics
of medicinal properties, sustainability in diversified environment
and soil conditions, less irrigation water requirement, high returns
on investment and huge export and domestic demand, pomegra-
nate is a good choice for cultivation. Thus, it is an important fruit
crop for ensuring the livelihood of farmers in water-scarce regions
(Bhandari, 2012). However, during last one and half decades,
pomegranate growers in India have been in dire straits because
of the severe outbreak of bacterial blight disease caused by Xan-
thomonas axonopodis pv. punicae, which has resulted in heavy yield
losses (Sharma et al., 2017).
The bacterial blight-causing pathogen in pomegranate, X. axo-
nopodis pv. punicae, was identified in 1952 (Hingorani and
Mehta, 1952) but was not considered a serious threat. Gradually,
the disease has become a serious threat and outbreaks have
reached an epiphytotic level in several parts of central India, result-
ing in yield loss both in terms of quality and quantity (Kumar et al.,
2011, Sharma et al., 2017, Sharma et al., 2010). Bacterial blight dis-
ease of pomegranate has been reported in countries such as Pak-
istan (Akhtar and Bhatti, 1992), South Africa (Petersen et al.,
2010), and Turkey (Icoz et al., 2014).
X. axonopodis pv. punicae is a gram-negative rod-shaped and
non-sporing bacterium. In general, the mode of infection of this
pathogen is via rain, irrigation water, infected planting material,
insect vectors, farm implements and humans. This pathogen also
spreads via air and enters the host through natural openings or
physical injuries to plants, the daily average temperature between
25 and 35 °C and RH > 30% favours the rapid blight spread. (Sharma
et al., 2017). The blight affects all above ground parts of the plant of
which the fruits are the most vulnerable (Sharma et al., 2010, Singh
et al., 2015). After inoculation, X. axonopodis pv. punicae infection
has three prominent stages in the host. The stage 1 is associated
with the appearance of water soaked lesions on leaves and fruit,
the stage 2 is associated with necrotic dark brown/blackish brown
spots with a yellow halo against light on leaf and necrotic dark
brown/blackish depressed lesions on fruit, and stage 3 is character-
ized by large blighted areas with dried silvery bacterial ooze on
leaves and enlarged necrotic lesions with cracks and dried white
encrustation of bacterial ooze on fruit (Sharma et al., 2017). PCR
based early and reliable diagnostic techniques are available for
detection and confirmation of X. axonopodis pv. punicae infection
in pomegranate (Sharma et al., 2017, Doddaraju et al., 2019).
Previously, transcriptome analysis based on two phenotypically
different pomegranate cultivars led to the identification of simple
sequence repeats (SSRs) and also single nucleotide polymorphisms
(SNPs) in pomegranate (Ophir et al., 2014). A few attempts have
been made to develop a genetic map of economically important
traits in pomegranate based on the transcript markers enriched
with quantitative trait loci (Harel-Beja et al., 2015). Whole genome
sequencing and limited transcriptome analysis using fruit peel,
leaves, roots, flowers along with their parts and fruit are available
to add to the large scale comparative transcriptome data generated
in the current investigation, all these can result in a tremendous
increase in genomics resources on pomegranate (Ono et al., 2011,
Ophir et al., 2014, Qin et al., 2017, Yuan et al., 2018).
Although these resources provide good information about the
biology of pomegranate, host–pathogen interactions with special
reference to X. axonopodis pv. punicae via RNA-Seq is not available.
To the best of our knowledge, this is the first report describing the
identification of the differentially expressed candidate genes
responsible for resistance against pathogen infection by using
NGS technologies such as RNA-Seq analysis. Here, we report results
from high-throughput Illumina sequencing, de novo assembly and
3515
functional annotation of the pomegranate transcriptome under
Xanthomonas infection and the identification of differentially
expressed genes (DEGs) in susceptible and moderately resistant
pomegranate genotypes. The transcriptome data generated could
help in understanding resistance and susceptibility nature of culti-
vars and further detailed fine mapping and functional validation of
identified candidate gene would provide scope for resistance
breeding programme in pomegranate.
2. Material and methods
2.1. Experimental site and climatic conditions
The study was carried out at Kegaon Research Farm, ICAR-
National Research Centre on Pomegranate, Solapur, India, having
17°430 N latitude, 75°500 E longitude and 486 m altitude from
mean sea level. During the experimental period, the minimum
average temperature and maximum average temperature of the
month ranged from 18.19 °C to 33.16 °C and the mean relative
humidity during the period was 66.87% with average wind velocity
6.37 ms1.
2.2. Plant material and challenge inoculation
’Bhagawa’ is India’s most commercial but blight susceptible
pomegranate cultivar having red rind and sweet soft red arils
where as IC 524207 is a wild type pomegranate genotype having
small acidic fruits but moderately resistant to blight. These pome-
granate plants were inoculated by spraying pure culture of X. axo-
nopodis pv. punicae under conditions favourable for blight infection
(average daily temperature 25-35 °C and RH > 30%). The culture
was maintained at the Plant Pathology Laboratory of ICAR-NRCP.
The pathogen was isolated from infected samples on NGA medium.
Briefly, samples were surface-sterilized in sodium hypochlorite
(2%) for 2–3 min, then washed with sterilized deionized water
three times and blot dried. The samples were macerated on a ster-
ilized glass slide and streaked on the NGA. The culture plates were
incubated at 28 ± 1 °C. The colonies that were developed after 48–
72 h of incubation were chosen for further confirmation. The colo-
nies of X. axonopodis pv. punicae were identified by colour, texture,
morphology and characteristic brown pigmented fuscan produc-
tion. The bacterial identity was further confirmed by gyrB-
specific PCR amplification that is known to yield a 491-bp ampli-
con in X. axonopodis pv. punicae (Mondal et al., 2012, Sharma
et al., 2017). To prepare the inoculum, a loop full of bacterium
was inoculated in nutrient glucose broth with constant shaking
at 100 rpm. Bacterial suspension was diluted to 10-7 with shaking
at 100 rpm. The spray inoculation method was used for challenge
inoculation under field conditions (Fig. 1B) and symptoms on leaf
and fruits were confirmed as bacterial blight infection by visual
identification (Fig. 1C–K) and in vitro culture characteristics of X.
axonopodis pv. punicae (Sharma et al., 2017).
2.3. Preparation of samples for RNA-Seq analysis
We used 12 samples (pooled total RNA samples from three
technical replicates) in this study: three leaves and fruit samples
of Bhagawa (pooled sample of three technical replicates) with pro-
gressive stages of blight infection symptoms from stage 1 to 3 were
collected, LS_1 i.e., first stage of infection on leaf of Bhagawa
appeared on 5th day after challenge inoculation (days to post
inoculation-dpi), second stage- LS_2 (8 dpi), third stage- LS_3 (11
dpi) and first stage on fruit of ’Bhagawa’-FS_1 (11dpi), second
stage-FS_2 (15 dpi), third stage-FS_3 (21 dpi), respectively; along-
with non-inoculated control leaf and fruit samples of Bhagawa
3516 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528

Fig. 1. (A) Stage-wise cultures of Xanthomonas axonopodis pv. punicae. (B) First stage of Xanthomonas infection on fruits of ‘Bhagawa’ at 11 dpi (susceptible variety) — water-
soaked lesions on leaves and appearance of oily patches (C) Second stage of infection at 15 dpi — appearance of necrotic dark brown/blackish brown depressed lesions. (D)
Third stage of infection at 21dpi — enlarged necrotic lesions with cracks and dried white encrustation of bacterial ooze on lesions. (E) First stage of infection on leaf of
Bhagawa at 5 dpi (F) Second stage on Bhagawa leaf at 8 dpi— necrotic dark brown/blackish brown spots with yellow halo against light. (G) Third stage on Bhagawa leaf at 11
dpi — large blighted areas with dried silvery bacterial ooze. (H) First stage of infection on fruit of IC524207 at 14 dpi.

(LS_C, FS_C); three pooled samples each of leaf and fruit of IC
524204 at stage 1 of blight infection was taken (LT_1: 7 dpi and
FT_1: 14 dpi) along with control samples of IC 524204 (LT_C,
FT_C). The samples of 2nd and 3rd stages of infection on both leaf
and fruit of IC 524204 were ignored as symptom progression was
very slow and overlapping infection symptoms were observed at
stage 1 and 2, however infection stage 3 (cracking of fruits) was
not observed on fruits of tolerant genotype. Total RNA was
extracted from all 12 samples by using the Nucleospin RNA isola-
tion kit. RNA quantity and quality were determined by using Qubit
Flurometer 3.0 with absorbance ratios at 260 nm/280 nm by using
Nanodrop. The RNA integrity was tested by separating an aliquot of
the RNA sample on 1% agarose gel electrophoresis. Quality was
assessed by visualizing and checking the appearance of ribosomal
RNA bands and by lack of degradation products. The integrity of
RNA was re-confirmed by using Bioanalyzer 2100 (Agilent, Folsom,
CA). Total RNA from three technical replicates were pooled for each
sample and samples were enriched for mRNA by using oligo (dT)
beads followed by removal of rRNA by using a Ribo-Zero kit. Then,
mRNA was fragmented randomly into small pieces by adding the
fragmentation buffer provided with the Illumina mRNA-seq kit
(Illumina, San Diego, CA). The fragmented mRNA was reverse-
transcribed by using random hexamers and further amplified to
produce double-stranded DNA. Repair of the fragment ends and
30 -end adenylation were implemented with the NEBNextTM DNA
Sample Prep Reagent Set 1 (New England BioLabs, Ipswich, MA).
The paired end adaptors were linked to the ends of the DNA frag-
ments. The resulting cDNA templates were gel-purified and PCR-
enriched. Quality of the mRNA-Seq library was verified by using
Bioanalyzer.
2.4. Illumina sequencing, quality control and transcriptome de novo
assembly
The quantified mRNA-Seq libraries were sequenced by using
Illumina HiSeq 2500. We used DrSeq, an automated RNA Seq pipe-
line developed by Nucleome Informatics, for data analysis. FastQC
was used to check the quality of the raw data and to avoid low-
quality bases that result in mis-assemblies at the time of assembly.
FastQC checks the quality of the data by considering sequencing
error rate (e) and sequencing base quality (Qphred) value or qual-
ity score (QC). The filtered high-quality reads were used for down-
stream analyses. The clean transcriptome reads were assembled de
novo by using Trinity software (Grabherr et al., 2011).
2.5. Functional annotation and classification of genes
Unigenes were searched against seven public databases: the
NCBI non-redundant protein sequence database (Nr database)
(https://www.ncbi.nlm.nih.gov/refseq/about/nonredundantpro-
teins/); the NCBI nucleotide sequence database (Nt database)
(https://www.ncbi.nlm.nih.gov/nucleotide?cmd=search); the KO
database (https://www.genome.jp/kegg/ko.html); SwissProt
(Boeckmann et al., 2003); Proteinfamily (Pfam) database
(Bateman et al., 2004); Geneontology (GO) database (Consortium,
2004) and Eukaryotic orthologous groups (KOG) database by using
BLASTX. The BLASTX results were imported to Blast2GO v2.5
(Conesa and Götz, 2008) for functional classification. We used in
house scripts for mapping, retrieving and allocation of GO terms
to genes. These genes were also annotated with unique enzyme
codes (ECs) and KEGG maps (http://www.genome.jp/kegg) by
using KAAS software (Kanehisa and Goto, 2000, Kanehisa et al.,
2011). The genes present in the transcriptome of bacterial blight-
challenged tissues were classified into three GO categories: biolog-
ical process, cellular component and molecular function.
2.6. Estimation of transcript abundance and analysis of differential
expression
The quantification of transcripts involved RNA-Seq with Expec-
tation Maximization RSEM 1.2.7 (https://www.encodeproject.
org/software/rsem/). To estimate the abundance of each samples
at various stages, the expression-normalizing fragments per kilo-
base of exon fragments per million mapped (FPKM) were calcu-
lated by using RSEM 1.2.7. P-values for differential expression of
genes among two samples was calculated by using the likelihood
ratio test in the R package, DEGseq (Wang et al., 2009; Marioni
et al., 2008). The P-values were adjusted by using the Q-value
(Storey and Tibshirani, 2003). Q-value < 0.05 was used to identify
the significantly differentially expressed genes. Finally, the combi-
nation of Q-value (<0.05 and fold change (>2 and < -2) was used to
identify up- and down regulated genes.
2.7. Prediction of R genes among the upregulated genes of different
sample combinations
We used all upregulated gene IDs for all combinations of
samples as described earlier. We downloaded the R gene protein
sequences of different plant species from (http://prgdb.crg.eu/
 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528 3517

wiki/Download) and a local database was constructed. By using the 3. Results

protein sequences obtained by translating the upregulated
unigenes of the combinations of the samples compared, BLASTP
search was performed between the protein sequences of upregu-
lated genes of all combinations and the local database of R
proteins.
2.8. Total RNA isolation and first-strand cDNA synthesis
Total RNA was isolated from pomegranate leaf and fruit tissues
of the 12 samples of Bhagawa and IC 524204 (with three technical
replicates) by using the NucleoSpin RNA plant kit (Macherey–
Nagel, Duren, Germany) according to the manufacturer’s protocol
with slight modifications. The RNA concentration and purity were
assessed by using the Qubit 3.0 fluorometer. The integrity was
checked on 1% (w/v) agarose gel. Total RNA samples (800 ng) each
with 260/280 ratio 2.0 to 2.1 and 260/230 ratio 2.1 to 2.3 were
used for cDNA synthesis for all samples. Total RNA from 3 technical
replicates of each sample were pooled together. First-strand cDNA
was synthesized by using a High-Capacity cDNA Reverse Transcrip-
tion kit (Applied Biosystems, USA), following the manufacturer’s
instructions, in a final volume of 20 ml. The final cDNA samples
were diluted three-fold and used as a template in RT-qPCR.
2.9. Quantitative real-time PCR (RT-qPCR)
The RT-qPCR reactions were performed with a Light Cycler 96
system (Applied Biosystems StepOne Plus, USA) with the Hi-SYBr
Master Mix including Taq polymerase (HiMEDIA, India) in 96-
well optical reaction Roche plates. Each reaction was performed
in triplicate with 5 ml SYBR Green Master, 1 ml template cDNA,
0.2 ml each primer (10 mM), and 3.6 ml RNase-free water with a total
volume of 10 ml. The RT-qPCR profile was 95 °C (2 min), 40 cycles of
95 °C (5 s), 60 °C (30 s) with fluorescent signal recording and 72 °C
for 30 s. The melting curve was obtained by using a high-resolution
melting profile performed after the last PCR cycle, 95 °C for 15 s
followed by a constant increase in the temperature from 65 °C
(15 s) to 95 °C (1 s). All samples in three technical replicates were
repeated to confirm results.
3.1. De novo assembly and clustering of transcriptome from infected
tissues of pomegranate
The pomegranate variety Bhagawa is cultivated type and highly
susceptible to X. axonopodis pv. punicae infection and the genotype
IC 524207 wild type collected from hills of Himalayas, is moder-
ately resistant to X. axonopodis pv. punicae infection. With a high-
throughput IlluminaHiSeq 2500 sequencing platform, we obtained
50,10,95,768 raw transcriptome reads from both susceptible Bha-
gawa and moderately-resistant IC 524207 from X. axonopodis pv.
punicae infected and control fruit and leaf samples, which
accounted for approximately 75.16 GB of sequenced data (NCBI
Accession No. PRJNA361285). Furthermore, raw reads were sub-
jected to quality control (Fast QC) to obtain clean transcriptome
reads of 48,58,02,868, which represented about 72.87 GB
(96.95%) of high-quality sequence data with a >92% quality check
score for all samples and 94% for most samples at Q30 with high
GC content (>47.50). The data on sequencing, quality of data and
error rates are in Table 1. All these filtered reads were used in tran-
scriptome assembly. The minimum, mean, median, and max length
as well as N50 and N90 value and total nucleotides sequenced
were 201, 1184, 642, 23878, 2081, 471, and 22,29,23,628, respec-
tively (Table 2). The transcriptome assembly in our study resulted
from 188,337 unigenes with maximum and minimum read lengths
of 23,878 and 200 bp, respectively, with an average size of assem-
bled unigene of 1184 bp and 34,626 unigenes >2 kb in size (Table 3,
Fig. 2A and B), which indicated increased coverage as well asthe
depth and high quality assembly of the sequenced transcriptome
data.
3.2. Similarity searches of pomegranate transcriptome
Functional annotation of unigenes was performed by using sim-
ilarity searches. About 1,60,683 of 1,88,337 transcripts (85.31%)
were annotated in at least in one of the selected databases.
Furthermore, about 21,606 transcripts (11.47%) were annotated
in all seven databases (Table 4) (Fig. 3A). The individual count of
total number of genes annotated against the selected databases

Table 1
RNA Seq data. LS_1, LS_2, LS_3, LS_C represent the leaf sample and control sample of Bhagawa infected with Xanthomonas axonopodispv. punicae, along with stage of infection.
FS_1, FS_2, FS_3 represent the infected fruit sample and control sample of Bhagawa along with stage of infection. LT_1, LT_C represents infected leaf sample of IC 524207 with first
stage of infection along with its control sample. FT_1, FT_C represent the infected fruit sample at first stage of infection and its control sample of IC 524207.

Sample Raw Reads Clean Reads Clean Bases Size (GB) Error % Q20% Q30% GC Content
LS_1 58,279,956 56,911,978 8.54 0.01 98.21 95.57 50.07
LS_2 64,644,180 63,438,302 9.52 0.01 98.23 95.59 49.32
LS_3 57,946,290 56,346,098 8.45 0.01 97.77 94.31 49.85
FS_1 20,630,104 19,691,646 2.95 0.01 97.35 93.47 54.63
FS_2 19,578,544 18,764,108 2.81 0.02 97.06 92.7 54.39
FS_3 21,913,568 21,063,454 3.16 0.02 97.01 92.57 54.29
LS_C 49,579,848 48,131,138 7.22 0.02 96.83 92.01 50.84
FS_C 49,733,888 47,679,996 7.15 0.01 97.94 94.85 49.9
LT_1 59,916,394 58,049,980 8.71 0.01 97.8 94.69 49.99
FT_1 23,247,992 22,516,058 3.38 0.01 97.56 93.84 47.5
LT_C 55,469,742 53,602,820 8.04 0.01 97.91 94.53 51.03
FT_C 20,155,262 19,607,290 2.94 0.01 97.83 94.21 48.23

Table 2
Data for transcripts. Minimum length represents the minimum length of the transcript. Mean length represents the average length of transcript. Maximum length represents the
transcript length with maximum base pairs.

Minimum length of Mean Length of Median Length of Maximum Length of N50 N90 Total
Transcript Transcript Transcript Transcript value value nucleotides
201 1184 642 23,878 2081 471 222,923,628
3518 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528

is shown in Table 4. BLASTX searches showed that about 51.6% of pomegranate with Eucalyptus. However, very small proportions of
annotated sequences belong to other species (many different spe- sequences were annotated with reference to Vitis (5.8%), Theo-
cies of plants mixed together), followed by 29.3% in Eucalyptus broma cacao (5.2%), Ricinus commuis (3.9%), Jatropha curcas (3.2%)
grandis (Fig. 3B), which shows the evolutionary relationship of and Gossypium raimndii (2.6%) (Fig. 3B).

Table 3
De novo transcript assembly and its length distribution. Unigene interval length represents the length of the unigenes.

Unigene interval length 200–500 bp 500–1 Kbp 1–2 kbp >2 kbp Maximum length Total
Total no. of unigenes 71,157 49,611 32,943 34,626 23,878 188,337

Fig. 2. (A) Distribution of unigenes based on length frequency. (B) Unigenes length distribution and their number.
 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528 3519

3.3. Functional classification of pomegranate unigenes based on GO,
KEGG and KOG
Pomegranate unigeneswith non-redundant annotations were
functionally annotated by Gene Ontology (GO) analysis for biolog-
ical process, molecular function and cellular component. From a
total of 90,485 pomegranate unigenes, 68,464 were assigned to
biological functions. In total, 78,107 unigenes were assigned to
molecular function and 44,414 to cellular components (Fig. 4A).
Functional classification of transcripts with respect to Kyoto Ency-
clopedia of Genes and Genomes (KEGG) showed most were under
the metabolism category and gene processing category (Fig. 4B).
The transcripts we obtained were aligned to the Eukaryotic orthol-
ogous groups (KOG) database to predict and classify their possible
functions. Transcripts falling under categories ‘‘post translational
modification”, ‘‘protein turnover”, and ‘‘chaperones” were highly
clustered (Fig. 4C).
3.4. Analysis of differentially expressed genes
FT_1 and FT_C was noticed in comparison to all other conditions as
reflected from both the FPKM density distribution and box plot
graphs (Fig. 5A and B). It was also interesting to note that fruit tis-
sues of tolerant genotypes displayed large number of genes
expressed at stage 1 compared to susceptible plant tissues at dif-
ferent stages. Further, FT_1 has higher expression compared to
control FT_C indicating induced activation of defense related genes
after infection at stage 1. In-order to find genes with similar
expression patterns under various experimental conditions hierar-
chical FPKM clustering analysis was performed using the log10
(FPKM + 1) values. By clustering genes with similar expression pat-
tern, it is possible to discern unknown functions of previously char-
acterized genes or the functions of unknown genes. Genes within
the same cluster exhibit the same trends in expression under dif-
ferent conditions. Heatmap showed many genes which were up
or down regulated among different conditions. We found higher
differential up and down regulation of genes in fruit tissues FT_1
and FT_C. However, higher differential down regulation of genes
in FS_1, 2 and 3 as compared to FS_C (Fig. 5C).

3.5. Identification of DEGs between control and infected fruit and leaf
In order to compare gene expression levels under different
samples of IC 524207 and Bhagawa
experimental conditions, an FPKM frequency distribution graph
and box plot analysis were performed. Both these graphs display
To understand the transcriptional response of the host at the
overall gene expression levels in test samples. The significant dif-
time of infection, we identified the DEGs between control and
ferences for mean gene expression levels with wide range between
infected samples. We selected the combinations of control and
infected samples shown in Table 5. In our analysis, we ignored
Table 4 infection stage 2nd and 3rd of leaf and fruits in IC 524507 as symp-
Number of annotated reads in different databases and their percentage annotation in tom progression was very slow and overlapping of infection symp-
each database considered in the analysis. About 85.1% of unigenes were annotated in toms was observed at stage 1 and 2, however infection stage 3 was
at least one database, with about 11.7% of genes annotated in all databases.
not observed on fruits and leaves of tolerant genotype.
Database used for annotation Number of Percentage of
annotated reads annotation
3.6. Identification of DEGs in leaf and fruit samples of control and
NR 94,535 50.19 infected tissues of IC 524207
NT 129,632 68.82
KO 39,558 21
SwissProt 124,999 66.36 We identified DEGs between the control and infected fruit sam-
PFAM 89,152 47.33 ples (FT_1vs FT_C) of IC 524207 at stage 1 (14 dpi). In infected fruit
GO 90,485 48.04 samples (FT_1 vs FT_C), about 937 and 1047 genes were up- and
KOG 54,102 28.72
downregulated, respectively (Table 6) (Supplementary dataset 1
Annotated in all databases 21,606 11.47
Annotated in at least one database 160,683 85.31 & 2). Similarly, a total of 622 and 593genes were up- and downreg-
Total genes 188,337 ulated in infected leaf samples versus the control (LT_1 vs LT_C) at
7 dpi (Table 6) (Supplementary dataset 3 & 4).

Fig. 3. (A) Venn diagram of annotation of unigenes in different databases. (B) Homologs of unigenes in other plant species. Pie chart shows that most unigenes have homologs
in Eucalyptus grandis.
3520 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528

Fig. 4. (A) Bar graph showing the distribution of unigene annotations in GO database. (B) Bar graph showing the distribution of unigene annotations in KEGG database. Most
of the genes fall under the carbohydrate metabolism pathway. (B) Bar graph showing the annotations of unigenes in KOG database. Most of the genes fall under general
functions.

Fig. 5. (A) FPKM density distribution. (B) FPKM box plot. (C) H_cluster heat map for FPKM cluster analysis of differentially expressed genes. Red denotes genes with high
expression levels and blue denotes genes with low expression levels. The color range from red to blue represents the log10 (FPKM + 1) value from large to small.

3.7. Identification of DEGs between leaf and fruit samples of control
and infected Bhagawa tissues
In the first stage of infection of fruit in Bhagawa, 717 and 2735
genes were up- and downregulated, respectively at 11 dpi (FS_1 vs
FS_C) (Table 6) (Supplementary dataset 5 & 6). In leaf infected sam-
ples, a total of 595 and 705 genes were up- and downregulated at 5
dpi (Table 6) (Supplementary dataset 7 & 8). In the second stage of
infection of Bhagawa fruit, about 779 genes were upregulated and
2740 downregulated with respect to the control at 15 dpi (Table 6)
(Supplementary dataset 9 & 10). Similarly, a total of 1101 and 1185
genes were up- and downregulated in the infected samples of leaf
with respect to the control at 8 dpi (Table 6) (Supplementary
dataset 11 & 12). In the third stage of pathogen infection, 795
 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528 3521

and 2656 genes were up- and downregulated in infected fruit sam- 3.8. Comparative study of the identification of up and down-regulated
ples at 21 dpi (Table 6) (Supplementary dataset 13 & 14) and 952 genes among infected samples of IC 524207 and Bhagawa
and 1143were up- and downregulated in infected leaf samples
with respect to the control at 11 dpi (Table 6) (Supplementary We compared infected fruit and leaf samples at different stages
dataset 15 & 16) (dpi) in IC 524207 and Bhagawa along with control to retrieve the
DEGs. The combinations for this analysis are shown in Table 7.
Table 8 represents the comprehensive information for the com-
pared samples and total number of DEGs at each stage of infection.
Table 5 At the first stage of fruit infection in Bhagawa and IC 524207, 2102
Combinations of the samples considered for understanding the differential expression and 4221 genes were up- and downregulated, respectively. In com-
of the genes between the control samples and infected samples. FT and LT represent paring the second stage of fruit of Bhagawa and the first stage of IC
the infected fruit and leaf samples of IC 524207, respectively. FS and LS represent the 524207, 2305 and 4447 genes were up-and downregulated. Simi-
infected fruit and leaf samples of Bhagawa, respectively. C represents the control.
larly, in comparing the Bhagawa third stage and IC 524207 first
Days to post inoculation Combination stage for fruit, 2347 and 4455 genes were up- and downregulated,
14 dpi FT_1 and FT_C respectively. In comparing infected leaf samples at the first stage of
7 dpi LT_1 and LT_C Bhagawa and stage 1 of IC 524207, 1011 and 1077 genes were up-
11 dpi FS_1 and FS_C and downregulated, respectively. Furthermore, 771 and 768 genes
15 dpi FS_2 and FS_C
were up- and downregulated in leaf samples at stage 2 of Bhagawa
21 dpi FS_3 and FS_C
5 dpi LS_1 and LS_C and stage 1 of IC 524207. Moreover, 1302 and 1240 genes were up-
8 dpi LS_2 and LS_C and downregulated in comparing stage 3 of Bhagawa and stage 1 of
11 dpi LS_3 and LS_C IC 524207.

3.9. Identification of up and down-regulated genes among infected leaf

and fruits samples of Bhagawa.

Table 6
Combination of samples compared and total number of up- and downregulated genes
While comparing leaf and fruit tissues of different progressive
in infected fruit and leaf samples of IC 524207 and Bhagawa pomegranate. FT and LT blight infection stages, the maximum number of differentially
represent the infected fruit and leaf samples of IC 524207. FS and LS represent the expressed genes were found in FS_1 vs FS_3 (399 downregulated
infected fruit and leaf sample of Bhagawa. C represents the control. and 358 upregulated) (Table 9 and 10). However, GO enrichment
Sample Stage of Combination of Total no. of Total no. of was higher in LS_1 vs LS_3. The most significantly enriched GO
No. infection sample upregulated downregulated term were related to catalytic activity and metabolic process
genes genes (Fig. 6). Major KEGG enriched pathways for DEGs were related to
1 First stage FT_1 and FT_C 937 1047 ribosome, photosynthesis, plant pathogen interaction, phenyl-
2 First stage LT_1 and LT_C 622 593 propanoid biosynthesis and plant hormone signal transduction.
3 First stage FS_1 and FS_C 717 2735
4 First stage LS_1 and LS_C 595 705
5 Second stage FS_2 and FS_C 779 2740 3.10. Identification of R genes among the upregulated genes of all
6 Second stage LS_2 and LS_C 1101 1185 combinations
7 Third stage FS_3 and FS_C 795 2656
8 Third stage LS_3 and LS_C 952 1143
We identified the R genes among upregulated genes of different
combinations (Table 11). The comparison of FS_C vs FT_C and FS_2
vs FT_1 had more numbers (about 175 and 140, respectively) of R
genes.
Table 7
Combinations of samples considered for understanding the differential expression of
the genes at each stage. IC 524207 represents the moderately resistant pomegranate Table 9
variety, L represents leaf sample, F represents fruit sample, S represents susceptible Identification of up and down-regulated genes among infected leaf and fruits samples
and T represents tolerant. of Bhagawa.

Sample No. Combination Sample No. Sample Details Upregulated genes Downregulated genes
1 FS_1 and FT_1 1 FS_1 vs FS_2 288 374
2 LS_1 and LT_1 2 FS_1 vs FS_3 399 358
3 FS_2 and FT_1 3 FS_2 vs FS_3 355 295
4 FS_3 and FT_1 4 LS_1 vs LS_2 296 269
5 LS_2 and LT_1 5 LS_1 vs LS_3 376 265
6 LS_3 and LT_1 6 LS_2 vs LS_3 330 223

Table 8
Combination of the samples compared and total number of up -and downregulated genes in infected fruit and leaf samples of IC 524207 and Bhagawa pomegranate between each
stage of infection.

Sample No. Sample combination Stage of infection of IC 524207 Stage of infection of Bhagawa Total upregulated genes Total downregulated genes
1 FS_1 and FT_1 First Stage First Stage 2102 4221
2 LS_1 and LT_1 First Stage First Stage 1011 1077
3 FS_2 and FT_1 First Stage Second Stage 2305 4447
4 LS_2 and LT_1 First Stage Second Stage 771 768
5 FS_3 and FT_1 First Stage Third Stage 2347 4455
6 LS_3 and LT_1 First Stage Third Stage 1302 1240
3522 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528

Table 10
Gene Ontology classification of significantly enriched up- and down-regulated genes among the different diseased leaf and fruit samples of ’Bhagawa’

Samples compared Gene Ontology classification of upregulated genes Gene Ontology classification of major downregulated genes
Biological Processes Cellular Processes Molecular Functions Biological Processes Cellular Processes Molecular Functions
FS_1 vs FS_2 73 43 83 85 45 98
FS_1 vs FS_3 68 37 76 96 49 108
FS_2 vs FS_3 67 33 75 75 38 84
LS_1 vs LS_2 122 71 129 145 69 147
LS_1 vs LS_3 126 61 132 178 89 188
LS_2 vs LS_3 103 50 114 125 81 143

Fig. 6. GO enrichment of up and downregulated genes of infected leaf and fruit transcriptome of Bhagawa.
 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528 3523

Table 11 cosylase, superoxide dismutase, and alcohol dehydrogenase. We

The predicted upregulated R genes in each combination. Among the compared selected these genes randomly from the data generated from DEG-
samples, FS_C vs FT_C had the highest number of R genes among the upregulated
genes.
seq analysis. We designed the forward and reverse primers for the
genes and amplified the genes in all 12 samples (Table 12). The
Sample Combination Total number of upregulated gene that codes for xyloglucan endotransglycosylase was slightly
gene matching R genes
over-expressed in infected leaf samples of Bhagawa at stage 1
FS_1 vs FS_C 72 and 3 of infection as compared with the control (Fig. 7A). Similarly,
FS_1 vs FT_1 115
FS_2 vs FS_C 70
the gene that codes for xyloglucan endotransglycosylase was
FS_2 vs FT_1 140 under-expressed in infected leaf samples of IC 524207 at stage 1
FS_3 vs FS_C 72 of infection versus the control. Xyloglucan endotransglycosylase
FS_3 vs FT_1 126 showed higher expression in infected fruit samples of Bhagawa
FS_C vs FT_C 175
at stage 2 of infection versus the control. We did not obtain any
FT_1 vs FT_C 83
LS_1 vs LS_C 54 data for infected fruit samples of Bhagawa at stage 1 and stage 3
LS_1 vs LT_1 93 or IC 524207 at stage 2 (Fig. 7A).
LS_2 vs LS_C 114 For the gene coding for alcohol dehydrogenase, the control sam-
LS_2 vs LT_1 76 ple of infected leaf of Bhagawa had higher expression than at any
LS_3 vs LS_C 75
of the stages (Fig. 7B). We did not obtain any data for the control
LS_3 vs LT_1 110
LS_C vs LT_C 73 infected leaf samples of IC 425207. Similarly, we did not obtain
LT_1 vs LT_C 49 any data from infected fruit samples of Bhagawa at stage 1 and 2
(11 and 15 dpi), but the expression of alcohol dehydrogenase
was similar at stage 3 of Bhagawa and its control in infected fruit
3.11. Experimental validation of DEGs samples.
For the gene coding for superoxide dismutase, we did not obtain
To validate the results of RNA-Seq, we used RT-qPCR to validate any data for infected fruit samples of Bhagawa at 11 dpi and 21 dpi,
the expression of three genes coding for xyloglucan endotransgly- but we found over-expression of the gene in infected fruit samples

Table 12
List of forward primer and reverse primers used for RT-qPCR of the selected genes.

S.No. Name of the gene Forward Primer Reverse Primer

1 Xyloglucan endo- transglycosylase CCTCCCACAGGGTAGAGTAG GAAAGGGTGACAGAGAACAGAG
2 Alcohol dehydrogenase GTGGTTTGTAGTTCCCGAAGA AAGAGGTGATTGCCGAGATG
3 Superoxide dismutase AATTCAGCCGTACACAGACC CACTGGACCAAACTCCATCATA

Fig. 7. (A) Bar graph representing the Relative Quantification of the gene coding for Xyloglucan endotransglycosylase. (B) Bar graph representing the Relative Quantification
of Superoxide dismutase and (C) Represents the Relative Quantification of Alcohol dehydrogenase. For all the figures on X-axis name of the sample is given and on Y-axis the
sample’s relative quantification is denoted. In the figures A, B and C few columns are missing where the RT PCR does not show any amplification.
3524 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528

Gene Ontology classification of most significantly enriched up- and down-regulated genes among the different samples compared. NA represents the status of regulation which is not significant as compared with the other processes of
of Bhagawa at stage 2 versus the control (15 dpi). However, we did

(94), Catalytic activity (344)

Catalytic activity (310) and
not obtain any data for IC 524207 at stage 1 of fruit infection

Gene Ontology classification of most significantly enriched downregulated genes

Molecular Functions and

hydrolase activity (138)

Oxidoreductase activity
(Fig. 7C).

Binding (1563)/protein

Catalytic activity (581)

Catalytic activity (588)

No. of genes involved

Binding (1541)

binding (753)
4. Discussion

In 1952, X. axonopodis pv. punicae was identified in India and

NA

NA
was found to be highly evolved and a host-specific pathogen
(Hingorani and Mehta, 1952). At the time of infection, the infected

Cell (318), cell part (318),

Cellular Processes and No.
leaves show early water-soaked lesions that later convert to necro-

and intracellular (230)

tic blighting. The infected fruits show coalesced water-soaked
lesions leading to necrosis and development of small cracks

of genes involved
(Petersen et al., 2010). As an immunological response towards
infection, the plant infected with the pathogen at different stages
of infection is expected to have a definite transcriptional response
(Song et al., 2019). Disease progression and appearance of symp-

NA

NA

NA

NA

NA

NA

NA
toms on leaf and fruit tissues of tolerant type was slow which con-

Biological Processes and

firm moderate resistance in IC 524207 against Xap and similar

No. of genes involved

disease reaction pattern was obtained in resistant and susceptible

Metabolic processes

Oxidation reduction
kiwifruit genotypes upon Pseudomonas syringae infection in kiwi-

processes (92)
fruit by Song, et al. (2019). Using RNA-Seq, we analyzed the tran-
scriptome of these pomegranate samples infected with X.
axonopodis pv. punicae at a large scale. To the best of our knowl-

(315)
edge, these data are the first with high quality in pomegranate as

NA

NA

NA

NA

NA

NA
compared with previous reports (Ophir et al., 2014). From the pre-

binding (116), cation binding (117)

liminary analysis, before considering the comparative analysis, we

Oxidoreductase (105), metal ion

found a change in signal transduction genes during pathogenesis
Molecular Functions and No. of

Catalytic activity (519) and ion

(Fig. 4A). Similar observations were found in soybean during Pero-
nospora manshurica infection (Dong et al., 2018), which may be due

Catalytic activity (457)

Oxido-reductase (103)

Oxidoreductase (124)

Oxidoreductase (121)
to stronger signals as a result of bacterial invasion. Also, Fig. 4B

Oxidoreductase (95)
shows high expression of genes involved in chaperone activities,
genes involved

which are required for degradation of misfolded proteins during

binding (362)
bacterial stress and also to increase other protein turnover to cope
with the stress condition.
Gene Ontology classification of most significantly enriched upregulated genes

After preliminary analysis, in-depth data analysis was per-

NA

formed in three different ways: comparing the transcriptome data

of the infected fruit and leaf samples with the controls, by compar-
Cellular Processes and
No. of genes involved

ison of progressive disease transcriptome data of Bhagawa leaf and

fruits and a comparative study among the infected samples of IC
524207 and Bhagawa.
GO terms obtained by the comparison of the IC 524207 infected
fruit samples with their controls (FT_1 vs FT_C) showed that the
GO. The number in the brackets represents the number of genes in that process.

NA

NA

NA

NA

NA

NA

NA

genes involved in single organism transport and single organism

localization biological processes were upregulated and those
Single organism transport (149) and single

Single organism Metabolic processes (197)

Single organism Metabolic processes (235)

Single organism metabolic processes (226)

involved in cell, cell part, intracellular processes were downregu-

Single organism metabolic process (187)

and oxidation–reduction process (110)

lated with infection (Table 13) (Fig. 8A, B).

Biological Processes and No. of genes

Oxidation reduction processes (102)

and Oxidation and Reduction (107)

Comparison of infected leaf samples of IC 524207 and its control

and Oxidation and reduction (130)
and oxidation and reduction (132)

(LT_1 vs LT_C) showed upregulation of genes involved in biological

process including oxidation-reduction and molecular functions
organism localization (150)

Metabolic processes (577)

Metabolic processes (490)

such as oxidoreductase (Table 13) (Fig. 9A, B). At the time of infec-
tion, there may be a biotic stimulus to cope with the oxidative stress
caused by pathogen infection, so genes involved in oxidation-
reduction were upregulated (Li et al., 2016, Liu et al., 2017). We
found no significant KEGG pathways in both comparisons.
involved

The comparison of infected fruit samples of Bhagawa and its

control (FS_1 vs FS_C)(FS_2 vs FS_C)(FS_3 vs FS_C) shows upregu-
lation of genes involved in oxidation and reduction processes,
which predominantly represents the biological process. Under
the molecular functions category, oxidoreductase was upregulated,
Samples compared

which may be due to the oxidative stress described earlier

(Table 13). Under the downregulated category, we did not find sig-
FT_1 vs FT_C
LT_1 vs LT_C

FS_1 vs FS_C

FS_2 vs FS_C

FS_3 vs FS_C

LS_1 vs LS_C

LS_2 vs LS_C

LS_3 vs LS_C

nificant GO-enriched terms except in FS_1 vs FS_C and FS_2 and

FS_C, which showed downregulated genes in the binding and pro-
Table 13

tein binding category. KEGG annotations in all the comparisons

showed that metabolic processes such as photosynthesis, oxidative
 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528 3525

phosphorylation, and protein processing in endoplasmic reticulum
were predominantly enriched and most of the upregulated genes
were under these categories (Table 14). In general, processes such
as photosynthesis in plants were downregulated, which is a
defense strategy adopted by the infected plant (Rojas et al.,
2014). However, in citrus, during infection with Candidatus liberib-
acter asiaticus, which results in haunglongbing or citrus green dis-
ease, (Martinelli et al., 2012) transcriptome analysis showed
upregulation of photosynthesis genes. Similar to citrus, when com-
paring infected fruit samples of Bhagawa to controls, photosynthe-
sis genes were upregulated. A similar consideration was made in
analyzing citrus green disease (Martinelli et al., 2012). More num-
ber of both up and down regulated genes were observed during
first stage of fruit infection in tolerant type (FT_1 vs. FT_C) as com-
pared to number of up regulated genes in FS_1 vs. FS_C, FS_2 and
FS_C vs. FS_3 vs. FS_C results suggest that more upregulated DEGs
are involved in tolerant type and there is rapid response of tolerant
type to pathogen infection as compared to susceptible type. Higher
number of downregulated genes in control compared susceptible
leaf and fruit samples at different stages of blight infection as com-
pared to leaf and fruit transcriptome of tolerant types suggest
downregulation of genes involved in pathways important for dis-
ease resistance like phenylpropanoid biosynthesis, which mainly
carry out synthesis of lignin and flavonoids responsible for resist-
ing pathogen infection. Similar results were obtained by Song,
et al. (2019) while carying DEG analysis of Pseudomonas syringae

Fig. 8. (A) GO annotation of upregulated genes between the sample FT_1 vs FT_C. (B) GO annotations of downregulated genes between FT_1 vs FT_C.Y-axis represents the
total number of genes in each category. X-axis represents different GO classifications of up- and downregulated genes.
3526 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528

Fig. 9. (A) GO annotation of upregulated genes between the sample LT_1 vs LT_C. (B) GO annotations of downregulated genes between LT_1 vs LT_C.Y-axis represents the
total number of genes in each category. X-axis represents different GO classification of up- and downregulated genes.

pv. actinidiae resistant and susceptible kiwifruit genotypes upon
challenge inoculation.
The most enriched GO terms were related to oxidation reduc-
tion and protein binding in differentially expressed genes
(Table 13). Thedownregulated genes were found to be under the
spliceosome and RNA transport category (FS_1 vs FS_C, FS_2 vs
FS_C, FS_3 vs FS_C) (Table 14). When the infected leaf samples of
Bhagawa were compared with the control (LS_1 vs LS_C) (LC_2
vs LS_C) (LS_3 vs LS_C), most of the upregulated genes were in
metabolic processes on GO analysis and downregulated genes
were predominantly involved in catalytic activity (Table 13). The
KEGG annotation showed that processes such as photosynthesis,
phenyl propanoid biosynthesis and amino sugar/nucleotide meta-
bolism were downregulated (Table 14). This finding agrees with
previously published literature in that for the invocation of the
plant defense system toward the invading pathogen, energy plays
a critical role for the differential expression of several genes and
pathways (Scheideler et al., 2002). The genes involved in photosyn-
thesis and chrolophyll biosynthesis were downregulated at the
time of invasion of virulent and avirulent pathogens (Denoux
et al., 2008, Rojas et al., 2014, Swarbrick et al., 2006, Truman
et al., 2006).
We found that when comparing LS_2 vs LS_C, along with the
pathways that were downregulated in stage 1, starch sugar meta-
bolism was downregulated. In sweet orange, Citrus sinensis, the
starch levels were increased in leaves infected with C. liberibacter
asiaticus, which causes citrus huanglongbing, by about 3 to 7 folds
as compared with controls. The cross-talk between the pathogen
and the plant may lead to the differential expression of starch
metabolism genes and their downregulation (Fan et al., 2010). Sim-
ilar phenomena might be exhibited by infected pomegranate,
which needs further investigation.
As described in Table 8, combinations were selected for the
comparative transcriptome analysis between infected leaf and fruit
 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528 3527

Table 14
Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways enrichment analysis of up- and downregulated genes among the different samples compared. NA represents the
status of regulation, which is not significant as compared with the other processes of GO. The number in the brackets represents the number of genes in that process.

Samples compared KEGG classification of upregulated genes and their
number
LT_1 vs LT_C Photosynthesis (15), terpenoid biosynthesis (11)
FT_1 vs FT_C Carbon metabolism (24), biosynthesis of aminoacid
(20), protein processing in endoplasmic reticulum (38)
FS_1 vs FS_C Photosynthesis (35), oxidative phosphorylation (35),
protein processing in endoplasmic reticulum (26)
FS_2 vs FS_C Photosynthesis (36), oxidative phosphorylation (29),
protein processing in endoplasmic reticulum (28)
FS_3 vs FS_C Photosynthesis (36), oxidative phosphorylation (31),
protein processing in endoplasmic reticulum (28)
LS_1 vs LS_C Photosynthesis (12), oxidative phosphorylation (14)
LS_2 vs LS_C Plant pathogen interaction (18)
LS_3 vs LS_C Phenylpropanoid biosynthesis (14), phenyl alanine
metabolism (13)
KEGG classification of downregulated genes and their number
Thiamine metabolism and gap junction (9)
Ribosome (60)
Spliceosome (59)
Spliceosome (63), RNA transport (40)
Spliceosome (63)
Photosynthesis (16), amino sugar and nucleotide sugar metabolism (15)
Starch and sucrose metabolism (27), photosynthesis (24), amino sugar and
nucleotide sugar metabolism (24), phenylpropanoid biosynthesis (18)
Amino sugar and nucleotide sugar metabolism (26), starch and sucrose
metabolism (23), phenylpropanoid biosynthesis (14), biosynthesis of aminoacid
(20)

samples between IC 524207 and Bhagawa. While comparing FS_1 Acknowledgement

vsFT_1 and FS_2 vs FT_1, KEGG annotation showed that the carbon
metabolism pathway was upregulated (Supplementary dataset All authors acknowledge the Indian Council of Agricultural
18). At the time of pathogen infection, generally carbon metabo- Research, New Delhi for financial support and Dr. Kishor Gaikwad,
lism is high because of the upregulation of associated genes, which Principal Scientist, ICAR-NIPB, New Delhi for his valuable sugges-
turns on the defense mechanisms by the activation of defense tions in drafting the manuscript.
genes and finally leads to the accumulation of H2O2 and salicylic
acid, etc. There is down regulation of pathways which play critical
Declaration of Competing Interest
role in defense response against pathogen like protein processing
in endoplasmic reticulum, flavanoid biosynthesis, phenyl-
The authors declare that they have no known competing finan-
propanoid biosynthesis, etc, when transcriptome of infected fruit
cial interests or personal relationships that could have appeared to
samples Bhagawa was compared with infected fruit transcriptome
influence the work reported in this paper.
of IC 524207, this might be one of the reasons for fast progression
of blight infection in susceptible tissues (Korner et al., 2015, Song
et al., 2019). The downregulation of the carbon metabolism path- Ethics approval and consent to participate
way in infected leaf samples of Bhagawa when compared with
infected leaf samples of IC 524207 and (LS_1 vs LT_1 and LS_2 vs None.
LT_1) (Supplementary dataset 17 and 18) may occur because of
downregulation of genes involved in photosynthesis, which alter Consent for publication
the carbohydrate metabolism and was also observed in Potato
infected with Rhizoctonia solani (Aliferis and Jabaji, 2012). Finally, All the authors have approved the final article.
some of the important genes that code for xyloglucan endotransg-
lycosylase, superoxide dismutase, and alcohol dehydrogenase were
validated by RT-qPCR. Availability of data and material

None.

5. Conclusions
We aimed to study and understand the defense response of
pomegranate against the X. axonopodis pv. punicae infection, which
is the sole cause of blight in pomegranate. The data from this large-
scale transcriptome study by using Illumina sequencing was of
high quality and the largest pomegranate transcriptome data avail-
able to date. Moreover, in-depth comparative analysis of RNA-Seq
data, especially the comparison between infected leaf samples of
Bhagawa with the control provides insights into the biological sig-
nificance of downregulation of genes involved in photosynthesis.
Similarly, comparison of infected fruit samples of Bhagawa with
infected fruit samples of IC 524207 showed how the upregulation
of the carbon metabolism pathway would provide resistance
against the infection. This report gives information on the defense
strategies adopted by the pomegranate against X. axonopodis pv.
punicae and also provides the preliminary information about the
differentially expressed genes between two pomegranate varieties
and their possible role in tolerance/susceptibility to infection.
Funding
The project was fully funded by Indian Council of Agricultural
Research, New Delhi, India under the Flagship Project, Integrated
Approach to Eradicate Pomegranate Bacterial Blight (project
code- IXX11105).
Authors’ contributions
Design of the study by NVS, RKP, SP and UKR. Drafting of the
manuscript and manuscript preparation by NVS, RS, DBK, RKP,
PP, BK, PVPSA, UKR, and DSB. RNA sequencing and data analysis
by NVRM, AT, PVPSA, HK and DSB. Sample preparation and
collection of experimental materials by NVS, SP, DMM and VRS.
Preparation of pure cultures of pathogen, challenge inoculation,
confirmation of Xanthomonas axonopodis pv. punicae infection by
JS, SP, NVS. DNA isolation, purification and quantification and qPCR
validation by SKP. RNA isolation and cDNA synthesis by NVS, SKP.
3528 N.V. Singh et al. / Saudi Journal of Biological Sciences 27 (2020) 3514–3528

Appendix A. Supplementary material Liu, Y., Zhang, Z., Fu, J., Wang, G., Wang, J., Liu, Y., 2017. Transcriptome analysis of
maize immature embryos reveals the roles of cysteine in improving
agrobacterium infection efficiency. Front. Plant Sci. 8, 1778.
Supplementary data to this article can be found online at Marioni, J.C., Mason, C.E., Mane, S.M., Stephens, M., Gilad, Y., 2008. RNA-seq: an
https://doi.org/10.1016/j.sjbs.2020.07.023. assessment of technical reproducibility and comparison with gene expression
arrays. Genome Res. 18, 1509–1517.
Martinelli, F., Uratsu, S.L., Albrecht, U., Reagan, R.L., Phu, M.L., Britton, M., Buffalo, V.,
Fass, J., Leicht, E., Zhao, W., 2012. Transcriptome profiling of citrus fruit
References response to huanglongbing disease. PLoS ONE 7, e38039.
MayuoniKirshinbaum, L., Porat, R., 2014. The flavor of pomegranate fruit: a review.
J. Sci. Food Agric. 94, 21–27.
Akhtar, M., Bhatti, M.R., 1992. Occurrence of bacterial leaf spot of pomegranate in
Mondal, K.K., Rajendran, T., Phaneendra, C., Mani, C., Sharma, J., Shukla, R., Verma,
Pakistan. Pakistan J. Agric. Res. 13, 95–97.
G., Kumar, R., Singh, D., Kumar, A., 2012. The reliable and rapid polymerase
Aliferis, K.A., Jabaji, S., 2012. FT-ICR/MS and GC-EI/MS metabolomics networking
chain reaction (PCR) diagnosis for Xanthomonas axonopodis pv. punicae in
unravels global potato sprout’s responses to Rhizoctonia solani infection. PLoS
pomegranate. African J. Microbiol. Res. 6, 5950–5956.
ONE 7, e42576.
Morton, J.F. 1987. Fruits of warm climates. JF Morton.Miami, FL. 33189 505 p. ISBN:
Bateman, A., Coin, L., Durbin, R., Finn, R.D., Hollich, V., GriffithsJones, S., Khanna, A.,
0-9610184-1-0.
Marshall, M., Moxon, S., Sonnhammer, E.L., 2004. The Pfam protein families
N Syed, D., Chamcheu, J.C., M Adhami, V., Mukhtar, H., 2013. Pomegranate extracts
database. Nucleic Acids Res. 32, D138–D141.
and cancer prevention: molecular and cellular activities. Anti-Cancer Agents
Bayazit, S., Caliskan, O., 2018. Morpho-pomological and Chemical Diversity of
Med. Chem. (Form. Curr. Med. Chem.-Anti-Cancer Agents) 13, 1149–1161.
Pomegranate Accessions Grown in Eastern Mediterranean Region of Turkey.
Ono, N.N., Britton, M.T., Fass, J.N., Nicolet, C.M., Lin, D., Tian, L., 2011. Exploring the
Bhandari, P.R., 2012. Pomegranate (Punica granatum L). Ancient seeds for modern
transcriptome landscape of pomegranate fruit peel for natural product
cure? Review of potential therapeutic applications. Int. J. Nutrit. Pharmacol.
biosynthetic gene and SSR marker discovery F. J. Integr. Plant Biol. 53, 800–813.
Neurol. Dis. 2, 171.
Ophir, R., Sherman, A., Rubinstein, M., Eshed, R., Schwager, M.S., Harel-Beja, R., Bar-
Boeckmann, B., Bairoch, A., Apweiler, R., Blatter, M.-C., Estreicher, A., Gasteiger, E.,
Ya’akov, I., Holland, D., 2014. Single-nucleotide polymorphism markers from
Martin, M.J., Michoud, K., O’donovan, C., Phan, I., 2003. The SWISS-PROT protein
de-novo assembly of the pomegranate transcriptome reveal germplasm genetic
knowledgebase and its supplement TrEMBL in 2003. Nucl. Acids Res. 31, 365–
diversity. PLoS ONE 9, e88998.
370.
Petersen, Y., Mansvelt, E., Venter, E., Langenhoven, W., 2010. Detection of
Conesa, A., Götz, S., 2008. Blast2GO: a comprehensive suite for functional analysis in
Xanthomonas axonopodis pv. punicae causing bacterial blight on
plant genomics. Int. J. Plant Genom. 2008.
pomegranate in South Africa. Australas. Plant Pathol. 39, 544–546.
Consortium, G.O., 2004. The Gene Ontology (GO) database and informatics resource.
Qin, G., Xu, C., Ming, R., Tang, H., Guyot, R., Kramer, E.M., Hu, Y., Yi, X., Qi, Y., Xu, X.,
Nucleic Acids Res. 32, D258–D261.
2017. The pomegranate (Punica granatum L.) genome and the genomics of
Denoux, C., Galletti, R., Mammarella, N., Gopalan, S., Werck, D., De Lorenzo, G.,
punicalagin biosynthesis. Plant J. 91, 1108–1128.
Ferrari, S., Ausubel, F.M., Dewdney, J., 2008. Activation of defense response
Rojas, C.M., Senthil-Kumar, M., Tzin, V., Mysore, K., 2014. Regulation of primary
pathways by OGs and Flg22 elicitors in Arabidopsis seedlings. Mole. Plant 1,
plant metabolism during plant-pathogen interactions and its contribution to
423–445.
plant defense. Front. Plant Sci. 5, 17.
Doddaraju, P., Kumar, P., Gunnaiah, R., et al., 2019. Reliable and early diagnosis of
Scheideler, M., Schlaich, N.L., Fellenberg, K., Beissbarth, T., Hauser, N.C., Vingron, M.,
bacterial blight in pomegranate caused by Xanthomonas axonopodis pv. punicae
Slusarenko, A.J., Hoheisel, J.D., 2002. Monitoring the switch from housekeeping
using sensitive PCR techniques. Sci. Rep. 9, 10097. https://doi.org/10.1038/
to pathogen defense metabolism in Arabidopsis thaliana using cDNA arrays. J.
s41598-019-46588-9.
Biol. Chem. 277, 10555–10561.
Dong, H., Shi, S., Zhang, C., Zhu, S., Li, M., Tan, J., Yu, Y., Lin, L., Jia, S., Wang, X., Wu, Y.,
Sharma, J., Sharma, K., Kumar, A., Mondal, K., Thalor, S., Maity, A., Gharate, R.,
Liu, Y., 2018. Transcriptomic analysis of genes in soybean in response to
Chinchure, S., Jadhav, V., 2017. Pomegranate bacterial blight: symptomatology
Peronospora manshurica infection. BMC Genom. 19, 366.
and rapid inoculation technique for Xanthomonas axonopodis pv. punicae. J. Plant
Fan, J., Chen, C., Brlansky, R., Gmitter Jr, F., Li, Z.G., 2010. Changes in carbohydrate
Pathol. 99, 109–119.
metabolism in Citrus sinensis infected with ‘Candidatus Liberibacter asiaticus’.
Sharma, K.K., Sharma, J., Jadhav, V.T., 2010. Status of bacterial blight of pomegranate
Plant. Pathol. 59, 1037–1043.
in India. Fruit Veg. Cereal Sci. Biotechnol. 4, 102–105.
Grabherr, M.G., Haas, B.J., Yassour, M., Levin, J.Z., Thompson, D.A., Amit, I., Adiconis,
Singh, N.V., Abburi, V.L., Ramajayam, D., Kumar, R., Chandra, R., Sharma, K.K.,
X., Fan, L., Raychowdhury, R., Zeng, Q., 2011. Full-length transcriptome
Sharma, J., Babu, K.D., Pal, R.K., Mundewadikar, D.M., 2015. Genetic diversity
assembly from RNA-Seq data without a reference genome. Nat. Biotechnol.
and association mapping of bacterial blight and other horticulturally important
29, 644.
traits with microsatellite markers in pomegranate from India. Mol. Genet.
Harel-Beja, R., Sherman, A., Rubinstein, M., Eshed, R., Bar-Ya’akov, I., Trainin, T.,
Genomics 290, 1393–1402.
Ophir, R., Holland, D., 2015. A novel genetic map of pomegranate based on
Soloklui, A.A., Ershadi, A., Fallahi, E., 2012. Evaluation of cold hardiness in seven
transcript markers enriched with QTLs for fruit quality traits. Tree Genet.
Iranian commercial pomegranate (Punica granatum L.) cultivars. HortScience 47,
Genom. 11, 109.
1821–1825.
Hingorani, M., Mehta, P., 1952. Bacterial leaf spot of pomegranate. Indian
Song, Y., Sun, L., Lin, M., Chen, J., Qi, X., Hu, Z., Fang, J., 2019. Comparative
Phytopathol. 5, 55–56.
transcriptome analysis of resistant and susceptible kiwifruits in response to
Icoz, S., Polat, I., Sulu, G., Yilmaz, M., Unlu, A., Soylu, S., Bozkurt, I., Baysal, Ö., 2014.
Pseudomonas syringae pv. Actinidiae during early infection. Plos One 14 (2),
First report of bacterial blight of pomegranate caused by Xanthomonas
e0211913. https://doi.org/10.1371/journal.pone.0211913.
axonopodis pv. punicae in Turkey. Plant Dis. 98, 1427.
Storey, J.D., Tibshirani, R., 2003. Statistical methods for identifying differentially
Kanehisa, M., Goto, S., 2000. KEGG: kyoto encyclopedia of genes and genomes.
expressed genes in DNA microarrays. Methods Mol. Biol. 224, 149–157.
Nucleic Acids Res. 28, 27–30.
Swarbrick, P.J., Schulze-Lefert, P., Scholes, J.D., 2006. Metabolic consequences of
Kanehisa, M., Goto, S., Sato, Y., Furumichi, M., Tanabe, M., 2011. KEGG for
susceptibility and resistance (race-specific and broad-spectrum) in barley
integration and interpretation of large-scale molecular data sets. Nucl. Acids
leaves challenged with powdery mildew. Plant, Cell Environ. 29, 1061–1076.
Res. 40, D109–D114.
Truman, W., De Zabala, M.T., Grant, M., 2006. Type III effectors orchestrate a
Korner, C.J., Du, X., Vollmer, M.E., Pajerowska-Mukhtar, K.M., 2015. Endoplasmic
complex interplay between transcriptional networks to modify basal defence
reticulum stres signaling in plant immunity-at the crossroasd of life and death.
responses during pathogenesis and resistance. Plant J. 46, 14–33.
Int. J. Mole. Sci. 16 (11), 26582–26598.
Wang, L., Feng, Z., Wang, X., Wang, X., Zhang, X., 2009. DEGseq: an R package for
Kumar, M.R., Wali, S., Benagi, V., Patil, H., Patil, S., 2011. Management of bacterial
identifying differentially expressed genes from RNA-seq data. Bioinformatics
blight of pomegranate through chemicals/antibiotics. Acta Hort (ISHS) 890,
26, 136–138.
481–484.
Yuan, Z., Fang, Y., Zhang, T., Fei, Z., Han, F., Liu, C., Liu, M., Xiao, W., Zhang, W., Wu, S.,
Li, J., Zhu, L., Hull, J.J., Liang, S., Daniell, H., Jin, S., Zhang, X., 2016. Transcriptome
2018. The pomegranate (Punica granatum L.) genome provides insights into fruit
analysis reveals a comprehensive insect resistance response mechanism in
quality and ovule developmental biology. Plant Biotechnol. J. 16 (7), 1363–
cotton to infestation by the phloem feeding insect Bemisia tabaci (whitefly).
1374. https://doi.org/10.1111/pbi.12875.
Plant Biotechnol. J. 14, 1956–1975.